During lectures, it’s very common to ask questions to the students. But only a comparatively small number of students can ever answer directly, and even then it’s often ‘the usual suspects’. What about students who don’t get a chance to respond, or who don’t wish to speak up in front of the whole class? It’s also common to ask students to discuss things with each other, but are they discussing productively – do they have enough guidance on what to discuss – and can you get a feel for what they talked about, en masse?
Peer Instruction [1] is a structured way to boost the effectiveness of in-class questions and student discussion. It provides a clear framework that guides the instructor and the students through a learning ‘episode’. As an approach, it is now well established and has some solid research evidence behind it [2]. The ideal Peer Instruction question doesn’t involve simple recall of facts, or the pedestrian application of a formula, but should require conceptual understanding, interpretation, or synthesis of ideas.
Figure 1 illustrates the basic structure of a Peer Instruction ‘episode’. The instructor begins by posing a question. Students then think about their answer individually. This part is very important: it might be tempting to skip it and go straight on to the discussion, but it’s essential that students marshal what they know and understand, and commit to a position before they hear from other people. Otherwise, they won’t be compelled to confront a possibly flawed understanding, or might just acquiesce to the loudest or most persuasive nearby voice.
Having decided on their response, the students vote. This is most effectively done with some technological aid such as Top Hat, but you can do it with coloured cards or a show of hands; it’s the technique that matters, not the technology. Figure 2 shows the responses from this first vote from one of my own classes, Physics 1B.
As shown in Figure 1, if most of the class gets the question correct then you can skip ahead, but if there is a lack of consensus (the typical ‘sweet spot’ is for something like 30-70% of students to be initially correct) then you invite the students to find some people near them who have a different answer, explain their reasoning to each other, try to reconcile any differences, and reach a conclusion.
I usually allow a couple of minutes for the discussion, but it’s fairly easy to tell from the tenor of the hubbub whether the students are still having productive discussion or whether it’s starting to peter out and it’s time for a re-vote. Figure 3 shows the same question as Figure 2 but with the post-discussion responses. As is typical – and with no further instructor input at all – more students are now getting the question correct, having had an opportunity to hear alternative thoughts and reflect on their own approach to the problem.
To finish off the episode, you might wish to open up a whole class discussion by inviting some students to talk about their thinking or approach to the problem. I usually show both before and after response graphs to the students at this point: it’s useful for the students to see how their responses fitted in to the class profile, and any changes that have occurred. But it’s crucial to wait until this point – if you show the first graph before the discussion, it skews what the students talk about: they tend to focus on why the most popular response is correct (even if it isn’t!) rather than fully exploring the problem. Either way, it’s important that the instructor rounds up with a clear explanation of the problem; this helps tie up any loose threads for students who are still not fully confident of their understanding.
Peer Instruction doesn’t have to be limited to questions with a single correct answer: it’s also useful for questions of interpretation, where different viewpoints need to be defended with reference to evidence or supporting arguments. Monash University have put together a very useful set of resources and perspectives on Peer Instruction in the Humanities.
Peer Instruction is a great way of introducing structured interactivity into lecture sessions: it helps students to make progress in the moment, and gives the instructor very useful feedback on what the class understands. Give it a try!
[1] Eric Mazur, Peer instruction: A User’s Manual, Addison-Wesley, 1997. [2] Catherine H Crouch and Eric Mazur, “Peer Instruction: Ten years of experience and results”, Am. J. Phys. 69 (9), 2001.
Ross Galloway is a Senior Teaching Development Officer in the School of Physics and Astronomy. He teaches on the undergraduate programmes in physics and also conducts pedagogic research as a member of the Edinburgh Physics Education Research group (EdPER). His research interests include the development of student problem solving skills, diagnostic testing, and flipped classroom pedagogies.
